Abstract P156: Association of Snp Rs7903146 in Tcf7l2 with Incident Type 2 Diabetes in Normal Weight and Overweight/obese Individuals; A Pooled Analysis
Background: Few studies have examined GWAS (Genome-Wide Association Study) SNP associations in subgroups because there is limited power in single studies. To better understand the association of GWAS SNPs with incident T2D in normal weight and overweight/obese individuals, we examined the association of rs7903146 in TCF7L2 with incident T2D in three pooled prospective cohorts. We hypothesize the association with rs7903146 will be stronger in normal weight T2D than in overweight T2D.
Methods: African Americans and Whites in the Atherosclerosis Risk in Communities (ARIC) study, Multi-Ethnic Study of Atherosclerosis (MESA), and Coronary Artery Risk Development in Young Adults (CARDIA) study who were free of diabetes at the first study visit and had rs7903146 genotyping were included in the analysis. Incident T2D was ascertained based on fasting glucose ≥ 126 mg/dl, or self-report of diabetes medication use at subsequent study visits. Normal weight T2D was designated in participants with BMI < 25 kg/m2 and overweight/obese T2D was designated in participants with BMI ≥ 25 kg/m2 at the time of T2D ascertainment. Multinomial logistic regression was used to calculate odds ratios for incident normal weight and overweight/obese T2D compared to no T2D by rs7903146 genotypes.
Results: In the pooled cohort of 18,037, over a follow-up time ranging from 7-21 years, there were 1857 incident cases of T2D: 179 normal weight and 1678 overweight/obese. rs7903146 allele frequencies and estimated odds ratios were similar in Africans Americans and Whites, so combined analyses are presented. Each copy of the T (minor) allele of rs7930146 was associated with 1.17 (95% C.I. =1.08-1.26, p=.00005) greater odds of incident T2D in the entire combined cohort. Each copy of the T allele was associated with a 1.29 greater odds of incident normal weight T2D (95% C.I. = 1.03-1.61, p = .025) and a 1.15 greater odds of incident overweight/obese T2D (95% C.I. = 1.06-1.25, p = .0004). A stronger association of rs7903146 with incident normal weight T2D was consistently observed across all three cohorts and both racial groups. However, the strengths of associations for the two weight groups were not significantly different (p > .05), although, in this exploratory analysis, we had only 12% power to detect a significant difference given the observed odds ratio between these two groups in multinomial analyses.
Discussion: In this analysis, the effect size for the association of SNP rs7903146 appeared stronger in normal weight than overweight/obese T2D but the OR were not significantly different; GWAS variants may have stronger effects on T2D risk in normal weight individuals. The effect size for rs7903146 was smaller than observed in large case-control diabetes GWAS, possibly due to our prospective study design and our exclusion of many earlier-onset cases, which were prevalent cases of diabetes at the MESA and ARIC baseline exams.
- © 2013 by American Heart Association, Inc.